Tools and Skills

• ArcGIS Pro
• Spatial Analysis
• Convex Hull Analysis
• Buffer Analysis
• Spatial Join
• Kernel Density Analysis
• Map Design
• Environmental Data Interpretation
• Collaborative Research

Project Goals

This project focused on a few main questions:

• Did reduced human mobility during the COVID-19 lockdown allow coyotes to expand into new urban areas?
• How did the spatial extent of coyote sightings change before, during, and after lockdown?
• Was reduced human activity associated with changes in where coyotes were observed?
• What can these patterns tell us about coexistence between wildlife and people in urban environments?

My Role

This project was completed in collaboration with one partner, and responsibilities were shared evenly across all stages of the analysis. I worked extensively on importing and cleaning our CSV datasets, converting them into spatial point layers in ArcGIS Pro, and helping conduct the spatial analyses used throughout the project. We also worked together on map design, interpretation of results, and presenting our findings in a clear and visually effective way.

Methods

We combined multiple datasets and spatial analysis techniques to investigate how coyote activity changed as human mobility declined during the pandemic.

Key Takeaways

A few major patterns emerged from the analysis:

• The spatial extent of coyote sightings expanded during the 2020 lockdown compared to 2019 and 2021
• Parking citations near coyote sightings dropped sharply in 2020, reflecting reduced human mobility
• Kernel density maps showed more widespread coyote activity in urban areas during lockdown
• Together, these findings suggest that reduced human presence coincided with broader and more visible coyote movement throughout Los Angeles

Deliverables

This project resulted in several final deliverables, which are linked below.

• View the final presentation (PDF)
• Read the written project report

Reflection

This project gave me hands-on experience using ArcGIS Pro to work with spatial data and apply analytical techniques to a real-world environmental question. It strengthened my skills in data preparation, spatial thinking, and map-based storytelling, while also showing me how geographic analysis can be used to better understand the relationship between people and wildlife in urban environments.

What I found especially valuable was being able to connect technical spatial methods to a broader question about coexistence and city planning. The project reinforced how changes in human behavior can shape ecological patterns in ways that become visible through careful spatial analysis.

Tools and Skills

• R
• ggplot2
• Exploratory Data Analysis
• Personal Data Collection
• Affective Visualization
• Visual Encoding
• Data Interpretation
• Reflective Design

Project Goals

This project focused on a few main questions:

• How does my screen time vary across different days of the week?
• Is there a relationship between class/study time and screen time?
• How do other contextual factors, such as sleep, social plans, commuting, or deadlines, shape my daily screen use?
• How can personal data be visualized in a way that captures both patterns and lived experience?

Methods

This project combined quantitative data analysis with more interpretive visual design.

library(tidyverse) #general use
library(DT) #package for making interactive tables
library(plotly) #package for making interactive figures

screentime <- read_csv("screentime.csv")

# Using package DT to create an interactive table
datatable(data = screentime)

Code-Based Visualizations

#finding most recent observation date
most_recent_date <- max(screentime$Date)

plot1_static <- ggplot(data = screentime, #starting with my dataframe
       aes(x = `Day of Week`, #x-axis: categorical variable
           y = `Screen Time (hours)`, #y-axis: response variable
           color = `Day of Week`, #coloring by day of week
           text = paste("Date:", Date, #for interactive plot points
                        "<br>Day:", `Day of Week`,
                        "<br>Screen Time:", `Screen Time (hours)`, "hours"))) + 
  geom_jitter(width = 0.15, #adding horizontal jitter
              alpha = 0.7, #making points slightly transparent
              size = 2) + #point size
  labs(
    x = "Day of Week",
    y = "Screen Time (hours)", #changing axes labels
    title = "Screen time by day of the week", #title summarizing main message,
    #subtitle for most recent observation
    subtitle = paste("Most recent observation:", most_recent_date)
  ) + 
  scale_color_manual( #changing colors from default
    values = c("Monday" = "firebrick",
               "Tuesday" = "darkorange",
               "Wednesday" = "yellow3",
               "Thursday" = "green3",
               "Friday" = "blue4",
               "Saturday" = "slateblue4",
               "Sunday" = "pink3")
  ) +
  theme_light() + #changing theme from default
  theme(legend.position = "none") #removing legend

# convert to interactive plotly
plot1_interactive <- ggplotly(plot1_static, tooltip = "text") |>
  layout(
    font = list(family = "Sans"),  #editing font for figure title and axes
    hoverlabel = list(
      font = list(
        family = "Sans", #hover text font
        size = 13, #hover text font size
        color = "#FFFFFF" #font color for hover text
      )
    )
  )

#display interactive version of the plot
plot1_interactive

plot2_static <- ggplot(
  data = screentime, #starting with my dataframe
  aes(
    x = `Class + Study time (hours)`, #x-axis: continuous predictor
    y = `Screen Time (hours)`, #y-axis: response variable
    #custom hover text for the interactive plot
    text = paste(
      "Date:", Date,
      "<br>Class + Study Time:", `Class + Study time (hours)`, "hours",
      "<br>Screen Time:", `Screen Time (hours)`, "hours"
    )
  )
) +
  geom_point(
    color = "darkorange", #changing color from default
    size = 2, #point size
    alpha = 0.8 #making points slightly transparent
  ) +
  #adding linear regression trend line to show overall relationship
  geom_smooth(
    method = "lm",
    se = TRUE, #showing confidence interval around regression line
    color = "#014f86", #custom color for trend line
    inherit.aes = FALSE, #preventing hover text from applying to regression line
    aes(
      x = `Class + Study time (hours)`,
      y = `Screen Time (hours)`
    )
  ) +
  labs(
    x = "Class + Study Time (hours)", #changing axes labels
    y = "Screen Time (hours)",
    title = "Relationship between class/study time and screen time", #title summarizing main message
    #subtitle showing most recent observation date
    subtitle = paste("Most recent observation:", most_recent_date)
  ) +
  theme_light() #changing theme from default


#using plotly package to make the figure interactive
plot2_interactive <- ggplotly(plot2_static, tooltip = "text") |>
  layout(
    font = list(family = "Sans"), #editing font for figure title and axes
    hoverlabel = list(
      font = list(
        family = "Sans",
        size = 13, #font size for hover text
        color = "#FFFFFF" #font color for hover text
      )
    )
  )

#display interactive version of the plot
plot2_interactive

Affective Visualization

Alongside the code-based plots, I created a hand-drawn affective visualization to represent the same dataset in a more symbolic and personal way. The project was inspired by Giorgia Lupi and Stefanie Posavec’s Week of Goodbyes, which uses hand-drawn marks to encode personal experiences. Rather than focusing only on statistical relationships, this piece emphasizes how daily habits feel when viewed together over time.

The visualization uses a custom visual language:

• Shape represents screen time level
  
• Fill pattern represents class/study time
  
• Outline style indicates social plans
  
• Red underline marks low sleep
  
• Red dot marks an exam or project due

This format allowed me to visualize my habits in a way that feels more reflective than a conventional chart, while still encoding several variables at once.

Key Takeaways

Several patterns emerged from this project:

• Screen time varies across days of the week, with some weekend days showing more variability.
• There appears to be a slight negative relationship between class/study time and screen time.
• Contextual factors such as sleep, deadlines, and social plans likely shape screen time in ways that are not fully captured by a simple regression line.
• Affective visualization can reveal patterns and experiences that traditional charts may not communicate as clearly.

Reflection

This project helped me think about data in two different ways: as something to analyze statistically, and as something to represent personally. The ggplot visualizations helped identify general trends, while the hand-drawn affective visualization captured the context and variability of daily life.

Tools and Skills

• R
• ggplot2
• dplyr
• Exploratory Data Analysis (EDA)
• ANOVA
• Simple Linear Regression
• Multiple Linear Regression
• Log Transformation
• Residual Analysis
• Model Selection

Project Goals

This project focused on a few main questions:

• Which diamond characteristics are most strongly associated with price?
• How well can diamond price be predicted using carat, cut, color, and depth?
• Do transformations improve model fit and better satisfy regression assumptions?
• Which combination of predictors produces the strongest and most interpretable final model?

My Role

In this project, I completed the full analysis independently in R. This included selecting a random sample from the dataset, creating visualizations, fitting and interpreting regression models, checking model assumptions, comparing adjusted values across models, and identifying the final best-performing model.

Methods

I combined visualization, statistical testing, and regression modeling to understand the drivers of diamond price.

Key Takeaways

A few major findings emerged from the analysis:

• Carat was the strongest predictor of diamond price
• Cut and color both improved the model and helped explain additional variation in price
• Depth did not meaningfully contribute to predicting price and was dropped from the final model
• Log-transforming price and carat substantially improved the regression assumptions and overall model fit
• The final model using log(carat), cut, and color explained about 94% of the variability in log price

View the full report

Reflection

This project strengthened my understanding of how regression modeling can be used not just to make predictions, but also to compare predictors, evaluate assumptions, and improve model performance through thoughtful transformations. It gave me more experience interpreting categorical and quantitative predictors together in a multiple regression setting, and helped me see how model diagnostics directly shape the decisions made during analysis.

What I found especially valuable was the process of refining the model rather than stopping at the first strong result. Seeing how the log transformation improved the residual patterns, and how model comparison helped justify dropping depth while keeping cut and color, made the final model feel much more intentional and statistically sound.

Tools and Skills

• Data Journalism
• Datawrapper
• Python
• Exploratory Data Analysis
• Public Records Research
• Interviewing
• Data Visualization
• Storytelling with Data

Project Goals

This project focused on a few main questions:

• How have UCSB housing prices changed over the past decade?
• How do costs compare between residence halls and university apartments?
• How do occupancy types (single, double, triple) affect overall cost?
• How have the cost differences between occupancies changed over time?
• What do these trends mean for students trying to find affordable housing?

My Role

As part of the Daily Nexus Data Section, I worked on analyzing housing price data and translating those findings into visuals and narrative reporting. My responsibilities included collecting and organizing housing data, performing exploratory analysis, building charts using Datawrapper, and contributing to the reporting and writing of the final article.

The project also involved gathering context through interviews with students and university officials and incorporating those perspectives alongside the quantitative findings.

Methods

This project combined traditional journalism with data analysis to explain long-term housing trends at UCSB.

Key Takeaways

Several important trends emerged from the analysis:

• UCSB housing prices have risen steadily over the past decade
• Residence hall costs have increased faster than university apartment costs
• Meal plan requirements contribute significantly to residence hall expenses
• The cost savings from choosing triple occupancy have narrowed over time
• Rising housing prices are pushing more students to consider off-campus options

Deliverables

This project resulted in a published article and accompanying data visualizations.

• Read the full article in the Daily Nexus

Visualizations

Reflection

This project strengthened my skills in data journalism and communicating quantitative findings to a broad audience. It required balancing statistical analysis with narrative storytelling, making sure that the data remained accurate while also being understandable to readers without a technical background.

Working on this article also showed me how data analysis can play a role in public conversations about affordability and student life. By combining data visualization with interviews and reporting, the project helped translate housing cost trends into a story that directly impacts the UCSB community.

Tools and Skills

• Google Sheets
• Data Cleaning
• Categorization
• Frequency Analysis
• Data Interpretation
• Datawrapper
• Data Visualization
• Advocacy-Focused Communication

Project Goals

This project focused on a few main questions:

• How many dining hall options were offered overall?
• How many of those options were vegan?
• Of the vegan options, how many could reasonably be considered nutritionally sufficient entrees?
• How often were those nutritious vegan entrees actually served throughout the academic year?
• How could these findings be communicated clearly to support campus dining advocacy?

My Role

In this project, I independently analyzed dining hall menu data to support a campus advocacy effort focused on vegan food access. I worked directly in Google Sheets to organize the data, count and compare subsets of menu items, and identify patterns in how frequently different vegan options appeared. I then created Datawrapper visuals to present the results in a way that would be easy to understand and useful for advocacy.

Methods

This project combined spreadsheet-based analysis with visual storytelling to highlight the difference between simply offering vegan items and offering vegan meals that are both nutritious and consistently available.

Key Takeaways

A few major patterns emerged from the analysis:

• Vegan options made up a minority of overall dining hall offerings
• Within the vegan subset, many items could not be considered nutritionally sufficient entrees
• Only a relatively small number of nutritious vegan entrees were served consistently throughout the year
• Looking only at the total count of vegan items overstated how accessible substantial vegan dining options actually were
• Clear visualizations made the findings more effective for advocacy and communication

Final Visuals

Reflection

This project showed me that even relatively simple tools like Google Sheets can be powerful when paired with a clear question and strong communication. It reinforced the importance of distinguishing between raw counts and meaningful access, especially in a policy or advocacy context where numbers can easily be misleading without proper interpretation.

What I found especially valuable was the challenge of turning spreadsheet analysis into a more persuasive visual story. The project strengthened my ability to work with structured data in a practical setting and to present findings in a way that supports real-world decision-making.

Tools and Skills

• Python
• Google Colab
• Exploratory Data Analysis (EDA)
• Sentiment Analysis
• Qualitative Analysis
• Data Visualization
• Dashboard Development
• Stakeholder Communication

Project Goals

This project focused on a few main questions:

• How were consumers reacting to the Beats Pill launch?
• What product features or qualities were mentioned most often?
• What themes emerged from survey responses and social media discussions?
• How could these findings be turned into useful product or marketing recommendations?

My Role

In this project, I worked on the full analysis pipeline from data preparation to final presentation. This included collecting and organizing data, analyzing sentiment and discussion themes, creating visualizations, and summarizing the results in a way that would be useful for non-technical stakeholders.

Methods

I combined both quantitative and qualitative approaches to better understand consumer responses.

Key Takeaways

A few major themes emerged from the analysis:

• Consumer reactions were shaped by a mix of product features, brand perception, and launch messaging
• Social listening data revealed recurring points of praise and criticism that complemented the survey findings
• Combining quantitative summaries with qualitative review made the recommendations more grounded and actionable
• Clear visual storytelling was important for communicating insights to stakeholders without a technical background

Sharing Insights

Deliverables

This project resulted in several final deliverables, which are linked below.

• View the final presentation (PDF)
• Open the Claude insights dashboard
• Read the project summary

Reflection

This project strengthened my skills in consumer analytics, storytelling, and stakeholder communication. It also gave me more experience working with messy real-world data and thinking carefully about how to turn analysis into recommendations that are both clear and useful.

What I found especially valuable was the balance between technical analysis and communication. Beyond identifying trends in the data, the project required me to think about how to present those findings in a consulting-style format that could support real decision-making.